**Marc Sabater 1, M. Luisa Garcia-Romeu 1,\*, Marina Vives-Mestres 2, Ines Ferrer <sup>1</sup> and Isabel Bagudanch <sup>1</sup>**


Received: 6 July 2018; Accepted: 6 August 2018; Published: 8 August 2018

**Abstract:** There has been increasing interest in the processes that enable part customization and small-batch production in recent years. The prosthetic sector, in which biocompatible materials are used, is one of the areas that requires these types of processes; Incremental Sheet Forming (ISF) technology can meet these requirements. However, the biocompatible thermoplastic polymers formed by this technology have not yet been tested. Hence, the aim of this paper is to cover this gap in our knowledge by analyzing the effects of process parameters on the ISF process with the aim of optimizing these parameters before the actual production of, in this case, customized prostheses. Tests with polycaprolactone (PCL) and ultra-high molecular weight polyethylene (UHMWPE) were performed. Maximum force, surface roughness and maximum depth were statistically analyzed by means of response surface methodology and survival analysis. Spindle speed and tool diameter were shown to be the most influential process parameters in terms of maximum forming force and surface roughness for both materials. In contrast, survival analysis applied to maximum depth showed a greater influence of tool diameter in PCL sheets and a greater influence of spindle speed in the case of UHMWPE.

**Keywords:** single; point; incremental; forming; thermoplastics; biocompatible; temperature; process; parameters

### **1. Introduction**

The paradigms of manufacturing have evolved from craft production to mass production and then from mass customization to what S. J. Hu [1] calls personalization or personalized production. There are several important concepts and technologies that have facilitated the development of personalized production, including open-architecture products, personalization design, on-demand manufacturing systems and cyber-physical systems.

In the context of on-demand manufacturing systems, Incremental Sheet Forming (ISF) emerged to meet the demand for rapid prototyping and small-batch production. The process consists of a sheet being formed by means of a round-tipped tool (or punch), which makes a series of small incremental deformations in the sheet on a predefined path that is governed by a numerical control. It is a simple process that can be applied in a number of different fields, ranging from the automotive and aeronautic sectors to architecture. However, it is struggling to find a place in industrial production beyond the various research efforts carried out in recent years. More recently, the biotechnology sector has been attracting most of these research efforts, which have involved a variety of raw materials (metals, such as titanium or polymers, such as polycaprolactone) as well as different variants of the ISF process (such as Single Point Incremental Forming, or SPIF, i.e., without the presence of a die or Two Point Incremental Forming, TPIF, which fully or partially employs a die).

Within the biotechnology sector, the manufacture of prostheses using ISF process involves two of the key elements of personalized production: design personalization, since customized geometry required is for each patient and on-demand manufacturing systems, since ISF technology makes small-batch production possible. The ISF process also offers certain flexibility since it does not require dedicated machinery that would entail heavy investment. In fact, an ISF production system can be adapted from one of the most common machines found in the workshops: a computer numerically controlled (CNC) milling machine fitted with a clamping system for the sheet. In addition, if a TPIF system is required, the die can be made of wood or resin. Hence, transforming a milling machine into an ISF system is a very affordable solution.

Initially, the raw materials being formed in ISF systems were metallic, such as aluminum alloys (AA1050 or AA3003, for example) and some steels (DC04 and AISI 304, among others). Such materials were widely used because of their good formability. Other metallic materials, such as magnesium alloys and titanium alloys [2] need to be heated to be formed incrementally, resulting in increased complexity and cost. Recently, however, some researchers have focused on polymeric materials and thermoplastics [3], because these have the advantage of being able to be formed at room temperature using the ISF process.

With regard to the process itself, a common concern for researchers from the traditional literature review for metallic and polymer materials is how the process parameters (step down, sheet thickness, tool diameter and wall angle, among others) affect the finished product; that is to say, how they affect the various specific response variables. The aim of such work is to establish the optimal combination of process parameters for achieving the desired effect on these different response variables. Three of the response variables that have garnered most attention in ISF studies are: (i) the maximum axial force (FZ max), to ensure that the maximum capacity is not exceeded, especially when a machining center has been adapted; (ii) the attained final roughness (Ra), which serves as an indicator of the quality of the finished product; and (iii) the maximum depth (of penetration) (Z) of the tool, which serves as an indicator of the material's maximum formability before any tearing, wrinkling or breakage occurs.

Aerens et al. [4], working with aluminum and steel alloys, estimated the steady state axial forming force (FZs) by means of an analytical model that accounts for tensile strength, initial sheet thickness, tool diameter, scallop height, and initial wall angle. The works of Li et al. [5,6] investigated further by seeking an efficient model for tangential force prediction whereas Bahloul et al. [7] focused on minimizing the sheet thinning rate and the tool (or punch) loads. Recently, Centeno et al. [8] shed light on the importance of spindle speed in considering the variation force for metallic materials—a factor which is even more important for polymeric materials [9–11].

Final roughness is another aspect that has captured attention. Recently, Liu et al. [12] aimed to provide a predictive model and the optimal process parameters for minimizing surface roughness in a study in which they incrementally formed a sheet of AA 7075 O-temper aluminum and investigated four process parameters: step down, feed rate, sheet thickness, and tool diameter. Radu and Cristea [13] discarded sheet thickness and introduced spindle speed for three materials: DC01 steel, 304 stainless steel and AA1050 aluminum alloy. Echrif and Hrairi [14] carried out a similar study on an AA1050-O aluminum alloy sheet.

Formability analysis, regardless of the material, is mainly focused on obtaining forming limit diagrams, e.g., Silva et al. [3], where dedicated equipment is required. However, since the initial stages of the development of ISF technology, alternative and simple formability indicators have been used, such as maximum reachable wall angle and its corresponding maximum reachable depth [2,15].

It is well known that using trial and error methodology for determining the best combination of process parameters in manufacturing processes is expensive and time consuming. Having reviewed the current status of ISF, we have noted that before addressing optimization, a Box-Behnken design (BBD) of response surface methodology can provide a systematic approach to examining the main effects of the process parameters—and the interactions between them—on the response variables. We have also observed that studies involving polymeric materials are scarce and, in the case of biocompatible materials, non-existent. Therefore, the aim of this paper is to investigate the effects of four process parameters: tool diameter, spindle speed, feed rate and step down on three response variables: forming axial force, surface roughness and final depth in the ISF process using two biocompatible thermoplastic materials: polycaprolactone (PCL) and ultra-high molecular weight polyethylene (UHMWPE).

#### **2. Materials and Methods**

#### *2.1. Geometry and Materials*

Tests were performed on 2 mm-thick sheets of the biocompatible polymers, UHMWPE and PCL. The test geometry in this experiment is a pyramidal frustum (Figure 1a), the features of which are:


**Figure 1.** (**a**) Test geometry (**b**) Dynamometer with tooling (**c**) Experimental setup (**d**) Roughness measurement on ultra-high molecular weight polyethylene (UHWMPE) (**e**) UHMWPE part (**f**) polycaprolactone (PCL) part.

From a general point of view, both polymers present low density and high ductility but differ in their thermal properties, e.g., Vicat and melting temperature are lower for PCL (Table 1). The commercially available UHMWPE sintered sheets were initially around 10 mm thick and were sliced by a CNC saw machine and converted into 150 × <sup>150</sup> × 2 mm<sup>3</sup> sheets. The PCL sheets were produced in our laboratory by compression molding. Around 55 g of PCL pellets (Sigma Aldrich, Saint Louis, MO, USA, ≈3 mm, average Mn = 80,000) were positioned into the cavity (150 × <sup>150</sup> × 2 mm3) of a stainless-steel cast which was previously warmed to a set temperature between 60 and 80 ◦C inside a heating hydraulic press. A low load was applied for a fixed time to guarantee the melt of the material. Subsequently, the load was increased, thus keeping the sheet in

place for a few more minutes to complete the final compaction of the fused polymer. The pressure was then retired, and the cast cooled to room temperature by placing it in a cooling press.


**Table 1.** Mechanical properties.

The selection of these two materials is made according to two aspects, their mechanical behavior and their final application. They represent, in both cases, two confronted or extreme cases. From the point of view of mechanical behavior, PCL shows a decrease in strength after the yield point, although it is maintained in a stable value. Whereas for UHMWPE a strain hardening behavior is appreciated (in [16] where self-made PCL and UHMWPE sheets are compared) demonstrating that is a more rigid material. While under the point of view of the final biomedical application, they also respond to two different possible sectors where ISF technology can develop products. The characteristic of biodegradability is very important for PCL; this is why at present it is highly valued by doctors. While the basic properties of UHMWPE have been significant in the orthopedic sector for years.
